Charging and SOC?

[PenderBen]

Hello all.

I did my best to search the forum on this topic but couldn’t find quite what I’m experiencing.
I have a Magnum inverter/charger (MS4448), Outback Flexmax 80 charger controllers (x2) and Simpliphi Phi 3.8 batteries (x3), and 8kw of panels (24 panels total).
The system is basically new, installed last summer on our new house and studio, it’s been running since then but we’re not in the buildings yet so it hasn’t really been in use other than occasional lighting and power tools. I didn’t design or install it; I’ve been learning the system since.

Originally the system only had 2 batteries, while going through various setups and issues I was having I came to realize there should be 3 batteries minimally (according to Simpliphi’s manuals), so I recently added the 3rd battery. One of the issues with 2 batteries was in the inverter‘s programming the minimum AHs you can set is 200AH, these are 75AH batteries; now I have 225AH, but in the programming it’s only adjustable by 10s, so I’ve set it at 220AH, not sure how this might be effecting things.
I went through the steps to properly add the new battery, got it up and running, reprogrammed the controllers and inverter, had no SOC showing on the remote (just ‘Think’n’), which I experienced before when something happened and the system died one night. The solution then was to charge with the generator until it registered a SOC. As I understand it it won’t register a SOC until it’s 100% charged, so that makes sense. I did have to run the generator for about 30 minutes after the batteries were effectively charged until it registered a SOC.

Here’s where I’m sort of running into problems; on a daily basis the SOC will get to 94% and the controllers will say ‘charged’. Often one controller will say ‘charged’ and the other will still be in some stage of charging, and they will change throughout the day, and I’ve seen the SOC creep up to as high as 98%, but typically not that high. Sometimes they both say ‘charged’, sometimes they will both start charging again. It’s possible some of this is influenced by the weather and they’re restarting due to varying light condition (it’s been overcast a lot), I’m not sure.
I realize the SOC isn’t the most accurate gauge, but prior to adding the 3rd battery it would register 100% regularly.
Does it sound like something is wrong? I was wondering if I need to somehow fully reset the inverter? I did a soft and hard reset while adding the new battery, but that didn’t seem to reset the programming or stats saved in the inverter; I didn’t set set via the remote.

All settings in the inverter and the controllers are set at whatever the manuals said they should be (basically whatever the Simpliphi manuals state), I haven’t tried adjusting any of those settings.

It would be nice to have an accurate SOC. Any advice?
Sorry for the long post… thanks for reading.

 

[sunshine_eggo]

What absorption voltage, duration and tail current on your FM80s?

Your SoC is probably pretty close to accurate.

The "charged" on the FM80 just means it's completed the charge cycle according to its programming. It doesn't actually mean the battery is charged.

 

[PenderBen]

Per Simpliphi’s specs:
absorption voltage- 56.4v
absorb time- .1 hour
float voltage- 54v
rebulk- 51.2v

I’m not sure about tail current, still learning and that’s not a setting I’ve come across (by that name anyway).

When it’s ‘charged’ according to the controller it’s often under 56v, but I’m not there observing it when it hits that point and may have dropped a little once the charging stopped, it’s usually between 53.something and 55.something.

 

[sunshine_eggo]

That's kinda atypical. Most recommend charge to 14.4 x4.

I think it's called absorption end amps on the FM80.

Absorption time of 0.1 hour is really short. I'd extend that to 0.5 hour.

 

[PenderBen]

Well, like I said, I don’t understand it well enough to deviate from the recommended settings (yet). I see in my original post I failed to mention they’re ‘48v’ batteries, not ’12’, I’m guessing that’s where you’re getting 14.4 from?

As for the absorb end amps, I would have to confirm it on each controller, but if I did it according to these directions (which I’m fairly certain I did) then it should be 4.5.

Here’s what I’m going off from Simpliphi:

 

[sunshine_eggo]

Well, like I said, I don’t understand it well enough to deviate from the recommended settings (yet). I see in my original post I failed to mention they’re ‘48v’

You didn't need to. An MS4448PAE is a 48V inverter.

batteries, not ’12’, I’m guessing that’s where you’re getting 14.4 from?

Yes, hence the x4, i.e., 14.4x4 = 57.6 - a more typical charge recommendation.

As for the absorb end amps, I would have to confirm it on each controller, but if I did it according to these directions (which I’m fairly certain I did) then it should be 4.5.

With 0.1h of absorption time, it's unlikely to ever hit the end-amps. 4.5Ah is correct.

Again, recommend increasing absorption time to 0.5h.

Essentially, with the charge settings they prescribe, it's not unexpected that your battery monitor doesn't sync to 100%. The 94-98% you see is likely within 2-3% of being accurate at a guess.

Your FM80 are feeding current to the battery, so the voltage they measure is a little higher than the actual battery voltage because of the resistance in the wires. They are going to measure absorption voltage before the battery is actually at absorption voltage. They will then drop to float 6 minutes later regardless of the end amps or the battery's actual voltage.

The longer absorption time will allow for the battery to actually hit the prescribed absorption voltage and allow additional absorption time for the current to drop to 4.5A.

 

[PenderBen]

That makes sense, I wonder why Simpliphi would have you set it that way? I guess ultimately it’s safe for the battery and would give it a potentially longer life?
They’ve got me scared to try much with their threats of voiding the warranty on every page ;-)
Perhaps I’ll give that a try and see over a few days how it does.
Thanks for your time and advice @sunshine_eggo

 

[sunshine_eggo]

That makes sense, I wonder why Simpliphi would have you set it that way? I guess ultimately it’s safe for the battery and would give it a potentially longer life?
They’ve got me scared to try much with their threats of voiding the warranty on every page ;-)
Perhaps I’ll give that a try and see over a few days how it does.
Thanks for your time and advice @sunshine_eggo

Their recommendations make a little more sense with an integrated system that accurately measures voltage and net current into the battery.

Most integrated systems use open circuit voltage (OCV) from a separate measuring device like a shunt. OCV is the voltage measured without the influence of current flowing in the measurement circuit. Flowing current encounters resistance in the wires, so that alters the voltage measurement.

Your Magnum system includes a shunt that feeds voltage/current data to the inverter, so it can calculate SoC but it can't communicate that to the FM80 since the systems are not compatible.

Your FM80 operate based only on what they see between them and the battery. You may be outputting 60A with your FM80s, but they have no way of knowing how much of that 60A is going to loads or being used to charge. You likely see a difference in Amps between what your shunt says and what your FM80's say.

In my Victron system, I have:

Shunt
Inverter
MPPT
GX device (integrates the components)

The shunt measures OCV and NET current into the battery. it then feeds that information to all components, i.e., the inverter and MPPT don't rely on the values they measure, they use what the shunt says.

The MPPT KNOWS that 30A of the 50A its measuring is being used to charge, and it KNOWS the battery voltage is actually a little lower than what it mesaures, so it bases it's behavior on the voltage and current provided by the shunt, not what it measures.

The Simpliphi batteries in my system would be charged more accurately and to a higher state of charge due to the more accurate measurements.

I'm confident that you're getting your battery very nearly fully charged, and the 94-98% values are reasonably accurate. Another means of indicating full charge is to investigate the settings for your Magnum shunt. You can likely alter them so that the shunt triggers 100% based on how you are charging the battery. You've already alterd the capacity, so there might be a way to alter the "charged" voltage.

Lastly, you can validate this issue for yourself:
1) Get a voltmeter and measure the battery voltage AT the terminals and then compare that to the voltages reported by the FM80s
2) compare the current reported by the shunt (on the Magnum display) vs. the sum of the currents reported by the two FM80. The FM80 will be at least a little higher.

 

[PenderBen]

Thanks again @sunshine_eggo, I’ll do those checks when I have a chance.

 

[PenderBen]

Took a couple readings yesterday; the SOC on the Magnum remote read 91% and 53.82v, the FM80s said ‘charged’ and read 54.1/2v, the batteries themselves were at 53.8.
Today the SOC read 90% and one FM80 was ‘charged’ at that point, the other ‘float’.

It seems like the SOC progressively drops over time even though the controllers are charging to what should be the same amount, I noticed that before adding the third battery, but if I charge with the generator it will (likely) charge to 100% according to the Magnum.
Why might it be dropping like that over time?
Essentially nothing is running on this system, a couple boilers in standby (maybe run every couple days), smoke alarms and the inverter itself.

I haven’t had time to mess with it more than that yet, maybe this weekend, I can certainly try extending the absorption time and see what a changes.

 

[sunshine_eggo]

Took a couple readings yesterday; the SOC on the Magnum remote read 91% and 53.82v, the FM80s said ‘charged’ and read 54.1/2v, the batteries themselves were at 53.8.
Today the SOC read 90% and one FM80 was ‘charged’ at that point, the other ‘float’.

53.8 and 53.82 at the inverter are both open circuit voltages, so they should be very close. The FM80s may be reading too high, and that's contributing to your problem. As I recall, in the "charged" mode, they aren't providing much current if any at all.

Recommend you repeat the check after sundown when the FM80 can't provide any current. There is a voltage calibration process where you can define an offset and make them read the same as the other meters. Advanced menu - "Vbatt Calibration"



Also worth triple checking that all terminals are nice and tight.

It seems like the SOC progressively drops over time even though the controllers are charging to what should be the same amount, I noticed that before adding the third battery,

Likely due to:
Voltage drop effect described above between the FM80 and battery
Potential calibration issue with high voltage reading on the FM80.
Battery monitor not syncing properly due to a 6 minute absorption period.

Assuming you have the Magnum BMK, the battery monitor can't be programmed for a set voltage. It determines full charge based on lead-acid behavior during absorption, i.e., it determines full charge based on a combination of the battery voltage stabilizing over time under reducing current. 6 minutes is likely not long enough for it to establish this behavior.

Additionally, if you have the BMK battery efficiency setting to "auto" or something less than 99%, it's going to do exactly as you see because it expects MORE Ah to be input vs. extracted.

but if I charge with the generator it will (likely) charge to 100% according to the Magnum.

"likely" indicates you're guessing. Have you actually done this?

Why might it be dropping like that over time?

See above.

Essentially nothing is running on this system, a couple boilers in standby (maybe run every couple days), smoke alarms and the inverter itself.

Clearly something is running, or we wouldn't be having this conversation ?

I haven’t had time to mess with it more than that yet, maybe this weekend, I can certainly try extending the absorption time and see what a changes.

Please do the following:

Check and adjust FM80 voltage accuracy when they can't charge (evening or with PV disconnected).
Set absorption to 30 minutes
Set BMK battery efficiency to 99%

I feel compelled to mention that 99.99% of this is a measurement issue. Your batteries are being consistently charged to the same level every day your FM80 read "charged" and/or "float." The only thing wrong is the way the battery monitor is computing SoC.

 

[PenderBen]

53.8 and 53.82 at the inverter are both open circuit voltages, so they should be very close. The FM80s may be reading too high, and that's contributing to your problem. As I recall, in the "charged" mode, they aren't providing much current if any at all.

Recommend you repeat the check after sundown when the FM80 can't provide any current. There is a voltage calibration process where you can define an offset and make them read the same as the other meters. Advanced menu - "Vbatt Calibration"

This is a good idea, I hadn’t considered this (since I wasn’t really aware of it ).
Should that be done at night too, when I check the resting voltage? Should I then set them to read exactly what the batteries read at their terminals at that time?

View attachment 134343

Also worth triple checking that all terminals are nice and tight.

Will check the terminals for sure, no reason not to.

Likely due to:
Voltage drop effect described above between the FM80 and battery
Potential calibration issue with high voltage reading on the FM80.
Battery monitor not syncing properly due to a 6 minute absorption period.

Assuming you have the Magnum BMK, the battery monitor can't be programmed for a set voltage. It determines full charge based on lead-acid behavior during absorption, i.e., it determines full charge based on a combination of the battery voltage stabilizing over time under reducing current. 6 minutes is likely not long enough for it to establish this behavior.

Yes, I’m using the Magnum BMK, there aren’t many adjustments to be made, but I’m sure I left it in its default setting which was likely ’auto’ something, will check.

Additionally, if you have the BMK battery efficiency setting to "auto" or something less than 99%, it's going to do exactly as you see because it expects MORE Ah to be input vs. extracted.




"likely" indicates you're guessing. Have you actually done this?

I say likely because that’s what I had to do to get the BMK to register a SOC in the first place, and it said 100% at that time. I did wonder if it was actually overcharging as it sat at a higher voltage for a while before actually showing a SOC. I don’t recall the exact voltage, I believe it was 57.something.
The inverter charger is set to CC/CV (per Simpliphi), it remained in CC mode at around 60amps at that voltage and then switched to CV, amps dropped to around 30 then slowly decrease over about 30min, at about 9a is when it started to display the SOC and was 100%, it didn’t say charged (on the Magnum remote) at that time, but I shut the gen down at that point. That was the last time I ran the generator, when I added the 3rd battery.

See above.



Clearly something is running, or we wouldn't be having this conversation ?

That‘s pretty much it for loads currently, prior to having the boilers on (they’re almost always in standby as they’re for radiant heating and aren’t running much at all) it was just the inverter (and a few incredibly minor loads like a few smoke alarms) and at that point it was drawing the batteries down (according to the Magnum SOC) 5-10% overnight depending on the time of year (night length).

Please do the following:

Check and adjust FM80 voltage accuracy when they can't charge (evening or with PV disconnected).
Set absorption to 30 minutes
Set BMK battery efficiency to 99%

I feel compelled to mention that 99.99% of this is a measurement issue. Your batteries are being consistently charged to the same level every day your FM80 read "charged" and/or "float." The only thing wrong is the way the battery monitor is computing SoC.

Will do on those steps.
It’s reassuring to know they’re probably charging as they should and it’s just a measurement issue, still want to correct it as the SOC is a nice gauge to be able to glance at on the fly.

Thanks for all your help, it’s most appreciate.

 

[sunshine_eggo]

This is a good idea, I hadn’t considered this (since I wasn’t really aware of it ).
Should that be done at night too, when I check the resting voltage? Should I then set them to read exactly what the batteries read at their terminals at that time?

Yes.

Will check the terminals for sure, no reason not to.

Connection problems can really get you. I've seen it a number of times and even been victim of it a couple. Still working to try and make it my #1 recommendation.

Yes, I’m using the Magnum BMK, there aren’t many adjustments to be made, but I’m sure I left it in its default setting which was likely ’auto’ something, will check.

99% will definitely help.

I say likely because that’s what I had to do to get the BMK to register a SOC in the first place, and it said 100% at that time. I did wonder if it was actually overcharging as it sat at a higher voltage for a while before actually showing a SOC. I don’t recall the exact voltage, I believe it was 57.something.

Makes sense. The BMK appears to be very limited in the ability to customize, and I did read something about needing AC charging for the initial sync.

The inverter charger is set to CC/CV (per Simpliphi), it remained in CC mode at around 60amps at that voltage and then switched to CV, amps dropped to around 30 then slowly decrease over about 30min, at about 9a is when it started to display the SOC and was 100%, it didn’t say charged (on the Magnum remote) at that time, but I shut the gen down at that point. That was the last time I ran the generator, when I added the 3rd battery.

There's that 30 minute absorption we've been talking about.

Will do on those steps.
It’s reassuring to know they’re probably charging as they should and it’s just a measurement issue, still want to correct it as the SOC is a nice gauge to be able to glance at on the fly.

Thanks for all your help, it’s most appreciate.

The chargers will behave very consistently, so the batteries are achieving the same state when "fully charged" from the FM80's perspective. Pretty sure recalibration, the longer absorption and 99% efficiency will re-tare the SoC to read 100% most days.

 

[PenderBen]

Ah hah, 30 min absorption, I see that now

Okay, I’ve made those adjustments we’ll see how it does now. I was there in the dark with a flashlight and didn’t check the tightness on the terminals, but I can do that in daylight tomorrow.

One thing I thought was a bit strange, I was there around 9, totally dark and mostly overcast and one of the FMs was ’Zzzzz’ but the other was still in ‘MPPT Float’, it showed some input but I can’t think how the PV could be producing anything at that point, or could it?

 

[sunshine_eggo]

53.4 and 53.6, respectively. 0.5A shouldn't cause that regargless of if it's suspicious.

What did the meter and shunt say for battery voltage?

 

[PenderBen]

The meter (and at the terminals) said 53.0 at that point.
I just thought it was odd the right controller wasn’t sleeping as well

 

[sunshine_eggo]

The meter (and at the terminals) said 53.0 at that point.
I just thought it was odd the right controller wasn’t sleeping as well

I've seen FM80 behave a little oddly, so I'm not surprised. My neighbors' do some weird stuff.

0.4-0.6V difference is pretty nuts. Both need to be calibrated to read the same voltage as the meter/shunt. Recommend you do it at night or preferrably with the PV disconnected.

 

[PenderBen]

I adjusted them last night to match, admittedly the one controller was in that float stage and I didn’t cut off the PV.
I’ll check it again tonight and throw the disconnects for good measure.

 

[sunshine_eggo]

I adjusted them last night to match, admittedly the one controller was in that float stage and I didn’t cut off the PV.
I’ll check it again tonight and throw the disconnects for good measure.

0.5A shouldn't make much difference @ 53V. I don't think it's real anyway given the conditions.

 

[PenderBen]

Well, pretty much the same result today, 89% SOC and ‘charged’ on the controllers, the volts matched up across components though. Checked the terminals, seem to be adequately tight.

I wonder if there’s some way to reset the SOC? Maybe just have to reset the whole Magnum, reprogram then charge with gen to get a new baseline?

Looking through the FM manual, I wonder how much ambient temperature could be effecting it? It’s been in the low 40sF where they are located, they are not equipped with battery temperature compensation; the Magnum has a battery temperature sensor, not that that helps in this case.
I do have a heater on a thermostatic plug in with them to keep it above freezing, but it doesn’t come on until mid 30s.

I’ll probably just leave it for a while and see what it does over some days.